fix(scrape): warn when a requested screenshot cannot be downloaded
Warn on stderr when a requested screenshot cannot be downloaded in scrape --all, instead of skipping it silently.
Problem. With scrape --all --screenshot (or --full-page-screenshot), if downloading the screenshot URL failed (non-OK HTTP status or a network error), the catch/!response.ok paths skipped it silently. The run ended "Completed: N/N succeeded" with no screenshot.png and no indication of why.
Fix. Write a one-line warning to stderr in both paths, naming the page URL and the reason (HTTP status or error message). The skip itself and the exit code are unchanged, since a missing screenshot was already treated as non-fatal.
Test. scrape-all-screenshot.test.ts drives handleAllScrapeCommand with a mocked map/client and a mocked fetch, covering the non-OK and the throwing case. Both fail on main and pass with the change; the rest of src/__tests__/commands (441 tests) passes.
This pull request was prepared with the assistance of an AI coding agent (Claude Code) and submitted by Alex Castillo, who is responsible for its content.
Summary by cubic
scrape --all now warns on stderr when a requested screenshot can't be downloaded, instead of silently skipping it.
The warning names the page URL and the reason (HTTP status or error message). The skip behavior and exit code are unchanged. Adds tests covering both the non-OK response and thrown-error paths.